Trace Analysis of Compounds SPE-LC/MS/MS. Sample Preparation Applications Scientist CSD NEMC 2012, Thursday August 9th

Transcription

1 Trace Analysis of Compounds in Potable Water by On-Line SPE-LC/MS/MS Joan Stevens, Ph.D. Sample Preparation Applications Scientist CSD NEMC 0, Thursday August 9th

2 Why the Interest in Emerging Contaminants (EC) and Pharmaceuticals and Personal Care Products (PPCPs) in Water? > 3,000 drugs and PPCPs Interest and fate of drugs, hormones Hormones affecting fertility, changes in gender Limited knowledge of treatability for all contaminants High public visibility, concern 8/9/0

3 Requirements for Emerging Contaminant Analysis Methods Selectivity/Specificity Sensitivity-MRL determination by LCMS/MS Target List QA/QC Ruggedness/matrix effects Cost 3 8/9/0

4 Innovative LC-MS/MS Approach to EC Analysis Cost effective screen for as many analytes as possible Reporting levels-l 5t to 0 ng/l /L(ppt) Known target analytes and identification from experts Accuracy and precision comparable to existing drinking water methods % recovery Use multiple ion transitions (MRM)-misidentification 4 8/9/0

5 Examples of EC Concerns Pesticides/Herbicides Sugar Substitutes 5 8/9/0

6 Sample Preparation Requirements: Basic Steps Collection device Collection of water sample 000 ml Shipping/cost Analysis 6 8/9/0

7 Sample Preparation Off-line SPE 000 ml transported t to the lab Trace enrich onto a cartridge or disk Elute with few mls solvent, dry, reconstitute Amenable LC, % organic solvent ml eluent loaded into autosampler vial -00 ul injected, UHPLC to HPLC UV/MS detection End result 98-99% prepared sample is wasted 7 8/9/0

8 Sample Preparation On-line SPE (trace enrichment-spe) 00% of the prepared sample is loaded d Volume can be <5 ml Combined with more sensitive detection (MS/MS) Logistics of getting sample to the lab is simplified 8 8/9/0

9 Setup of the Agilent AQUA Online-SPE System SPE system: Quaternary pump G39A autosampler with 900 µl head 6port pos valve port 6 pos valve port stream selection valve (optional) RRLC system: Binary pump SL Well plate sampler SL (not used) Column department SL 9 8/9/0

10 Method Implementation in Mass Hunter Acquisition Single 900 µl injection: DRAW def. amount from sample VALVE mainpass WAIT 0.7 min /.5 min** REMOTE startpulse ** The delay volume of the online-spe system is approximately 00 µl. If the 900 µl seat extension is used the wait time has to be extended to.5 min. 0 8/9/0

11 On-line SPE method Autosampler Program(s) Single 900 µl injection: DRAW def. amount from sample VALVE mainpass WAIT 0.7 min /.5 min** REMOTE startpulse ** The delay volume of the system until the cartridge is approximately 00 µl. If the 900 µl seat extension is used the wait time has to be extended to.5 min (flush volume 3 ml). -fold 900 µl injection using the multiple draw kit (900 µl seat extension): DRAW def. amount from sample EJECT def. amount into seat DRAW def. amount from sample VALVE mainpass WAIT.5 min REMOTE startpulse 8/9/0

12 Online-SPE Method Loading Pump Mobile phase: A: 0.05% formic acid B: acetonitrile C: isopropanol SPE cartridge**: PLRP-s available. x.5mm or 4.6 x.5mm Temperature: Flow (load): ambient ml/min Loading and equilibration: 0.00 min 0 % B 0 % C.0 ml/min 0.0 min 0 % B 0 % C 0. ml/min.50 min 0 % B 0 % C 0. ml/min.60 min 0 % B 00 % C.0 ml/min.70 min 0 % B 00 % C.0 ml/min 4.00 min 0 % B 00 % C.0 ml/min 4.0 min 00 % B 0 % C.0 ml/min 5.50 min 00 % B 0 % C.0 ml/min 5.60 min 0 % B 0 % C.0 ml/min ** SPE cartridges fit in the Agilent Guard Column Hardware Kit (P.N ) All connections between the analytical system and the SPE hardware have been made by using 0.8 mm ID PEEK tubing (P.N ) and finger-tight PEEK fittings (P.N ). 8/9/0

13 Online-SPE Timetable for Loading and Analysis SPE IPA AcN 0.05% FA Cartridge equilibration HPLC separation of enriched pesticides min Equilibration of SPE cartridges, SPE and separation of pesticides is overlapped and results in a total runtime for the method of 0 minutes. 3

14 Example :High Sensitivity Analysis of Pesticides Atrazine Carbendazim Carbetamid Chloridazon Chlorotoluron Chloroxuron Desethylatrazine Desisopropylatrazine Desmetryn Diflubenzuron Diuron Fenuron Irgarol 05 Isoproturon Linuron Metamitron Methabenzthiazuron Metoxuron Metsulfuron methyl Monolinuron Monuron Neburon Prometryn Propazin Simazine Terbutryn Terbutylazine Ti Trietazin i 4

15 Example :High Sensitivity Analysis of Pesticides AQUA Online-SPE system consisting of Quarternary pump G39A autosampler with 900 µl head 6 port pos valve port 6 pos valve UHPLC system consisting of 90 Infinity binary pump 90 Infinity autosampler (not used) 90 Infinity it thermostatted t tt column compartment G6460AA Triple Quadrupole Mass Spectrometer t 5 8/9/0

16 Example :High Sensitivity Analysis of Pesticides Isoproturon - 9 Levels, 7 Levels Used, 8 Points, 4 Points Used, 0 QCs x0 y =.366 * x R^ = Responses Relative EIC for Cal_ (.0 ng/l) + MRM (07. -> 7.0) Cal_-r00.d x Isoproturon Relative Concentration Counts > 7.0, 07. -> 46. x0 3 Ratio=73..8 Counts Quantifier: Qualifier: Peak width: 6. s Data points: 8 Linear range: < ng/l to µg/l 6 8/9/0

17 Example : High Sensitivity Analysis of Pesticides x0 +ESI MRM Frag=0.0V CID@6.0 ( > ) BlankWater--r00.d Noise (PeakToPeak) = 8.3; SNR (7.473min) = x ESI MRM Frag=0.0V CID@6.0 ( > ) TestSample_-r00.d Noise (PeakToPeak) = 4.8; SNR (7.50min) = x0 3 x0 4 +ESI MRM Frag=0.0V CID@6.0 ( > ) Cal_-r00.d Noise (PeakToPeak) = 47.74; SNR (7.54min) = ESI MRM Frag=0.0V CID@6.0 ( > ) Cal_-r00.d Noise (PeakToPeak) = 69.74; SNR (7.490min) = x ESI MRM Frag=0.0V CID@6.0 ( > ) TestSample_-r00.d Noise (PeakToPeak) =.4; SNR (7.56min) = x0 4 x0 5 +ESI MRM Frag=0.0V CID@6.0 ( > ) Cal_3-r00.d Noise (PeakToPeak) = 79.88; SNR (7.487min) = ESI MRM Frag=0.0V CID@6.0 ( > ) Cal_4-r00.d Noise (PeakToPeak) = 80.6; SNR (7.5min) = x ESI MRM Frag=0.0V CID@6.0 ( > ) TestSample_3-r00.d Noise (PeakToPeak) = 0.96; SNR (7.58min) = Counts vs. Acquisition Time (min) Counts vs. Acquisition Time (min) 7 8/9/0

18 Example : High Sensitivity Analysis of Pesticides Cyanazine - 9 Levels, 8 Levels Used, 8 Points, 6 Points Used, 0 QCs + MRM (4. -> 4.0) Cal_-r00.d 4. -> 4.0, 4. -> 68.0 y =.8030 * x E-004 x0 9 R^ = x0 Ratio= Cyanazine EIC for Cal_ (.0 ng/l) Cyanazine Relative Concentration Responses Relative Counts Counts Quantifier: Qualifier: Peak width: 6.0 s Data points: 8 Linear range: ng/l to 5 µg/l 8 8/9/0

19 Example : High Sensitivity Analysis of Pesticides x0 +ESI MRM Frag=5.0V CID@** ( > ) BlankWater--r00.d Noise (PeakToPeak) = 9.4 x ESI MRM Frag=5.0V CID@** ( > ) TestSample_-r00.d Noise (PeakToPeak) = 40.6; SNR (6.643min) = x0 5 x0 3 +ESI MRM Frag=5.0V CID@** ( > ) Cal_-r00.d Noise (PeakToPeak) = 30.08; SNR (6.659min) = ESI MRM Frag=5.0V CID@** ( > ) Cal_-r00.d Noise (PeakToPeak) = 30.80; SNR (6.63min) = x0.5 +ESI MRM Frag=5.0V CID@** ( > ) TestSample_-r00.d Noise (PeakToPeak) = 3.58 x0 3 x0 4 +ESI MRM Frag=5.0V CID@** ( > ) Cal_3-r00.d Noise (PeakToPeak) = 34.38; SNR (6.68min) = ESI MRM Frag=5.0V CID@** ( > ) Cal_4-r00.d Noise (PeakToPeak) = 3.4; SNR (6.658min) = x ESI MRM Frag=5.0V CID@** ( > ) TestSample_3-r00.d Noise (PeakToPeak) = 49.56; SNR (6.648min) = Counts vs. Acquisition Time (min) Counts vs. Acquisition Time (min) 9 8/9/0

20 Example : High Sensitivity Analysis of Pesticides Bentazone - 9 Levels, 5 Levels Used, 8 Points, 0 Points Used, 0 QCs - MRM (39. -> 3.0) Cal_-r00.d 39. -> 3.0, 39. -> 97.0 x0 - y = 8.06 * x x x0 3 Ratio=30. R^ = Bentazone EIC for Cal_ (.0 ng/l) Relative Concentration Responses Relative Counts Counts Quantifier: Qualifier: Peak width: 9.8 s Data points: 3 Linear range: << ng/l to 0. µg/l 0 8/9/0

21 Example : High Sensitivity Analysis of Pesticides x0 x0 3 x0 4 -ESI MRM ( > ) BlankWater--r00.d Noise (PeakToPeak) = 3.40; SNR (7.38min) = ESI MRM ( > ) Cal_-r00.d Noise (PeakToPeak) = 8.30; SNR (7.389min) = 7.6 x ESI MRM ( > ) Cal_-r00.d Noise (PeakToPeak) = 6.56; SNR (7.37min) = x ESI MRM ( > ) TestSample_-r00.d Noise (PeakToPeak) = 66.44; SNR (7.39min) = ESI MRM ( > ) TestSample_-r00.d Noise (PeakToPeak) = 6.84; SNR (7.393min) = x0 4 x0 5 -ESI MRM ( > ) Cal_3-r00.d Noise (PeakToPeak) = 39.70; SNR (7.367min) = ESI MRM ( > ) Cal_4-r00.d Noise (PeakToPeak) = 83.36; SNR (7.400min) = x ESI MRM ( > ) TestSample_3-r00.d Noise (PeakToPeak) = 9.50; SNR (7.387min) = Counts vs. Acquisition Time (min) Counts vs. Acquisition Time (min) 8/9/0

22 Example : High sensitivity analysis of pesticides Compound Average (µg/l) OH Atrazine Atrazine desisopropyl ,6 Dichlorobenzamide Metamitron Atrazine desethyl th Dimethoate Hexazinone Simazine Cyanazine Bentazone Atrazine SD RSD Compound Average (µg/l) SD RSD Isoproturon ,4 D MCPA DNOC Mechlorprop Dichlorprop Terbuthylazin Dinoseb Pendimethalin

23 Example : Impact of Solvent and Sample Composition In addition to method for the neutral herbicides an on-line SPE method for the following 5 compounds (LOQ < 0 ng/l): Flufenacet Iodsulfuron-methyl Isoxaflutole Mesosulfuron-methyl Mesotrione Sample solution contains 80 mg/l Na SO 4 for sample conservation Compounds belong to group of acidic herbicides, Mesotrione shows broad peak in chromatography 3

24 Example : Impact of Solvent and Sample Composition Calibration standard 5 (0 to 97 ng/l); injection volume 900 µl x0 4 +ESI MRM Frag=80.0V CID@** ( > ) Cal_00ppt_900ul-r00.d one Mesotri methyl Me esosulfuron Io odosulfuron mehtyl Is soxaflutole Flufenacet Counts vs. Acquisition Time (min) 4 8/9/0

25 Example : Impact of Solvent and Sample Composition Flow rate of the loading pump had no influence on loading efficiency and therefore a flow of ml/min was chosen to reduce loading time. Compared to the use of demineralized water as the loading solvent loading with 0.% formic acid increased the response significantly Loading solvent 0.% formic acid Loading solvent demineralized water Peak areas in counts Mesotrione Mesosulfuron methyl Iodsulfuron methyl Isoxaflutole Flufenacet For injection volumes > 900 µl the addition of 0.% formic acid to the sample solution is strongly recommended. 5 8/9/0

26 Example : Impact of Solvent and Sample Composition The effect of the addition of acetonitrile and Na SO 4 (80 mg/l) to the sample solution on the on-line SPE recovery was tested Acetonitrile addition Tap water NaSO4 addition counts Peak areas in Mesotrione Mesosulfuron methyl Iodsulfuron methyl Isoxaflutole Flufenacet Acetonitrile addition was detrimental, the addition of Na SO 4 increased the mesosulfuronemethyl response. It is recommended to prepare samples and calibration standards identical. 6 8/9/0

27 Example 3: Robust Method for Glyphosate and AMPA Glyphosate is a global herbicide which is widely used in agriculture and urban landscape management In the environment glyphosate is metabolized to its metabolite aminomethyl phosphonic acid (AMPA) Both compounds are extremely polar due to their bipolar structure For the analysis of both compounds a derivatization is widely accepted 7

28 Example 3: Robust Method for Glyphosate and AMPA Derivatization works at basic ph-values with high excess of FMOC and both reduces the lifetime of the chromatographic column and the robustness of the method Sensitivity of direct injection is just enough to detect 00 ng/l Online-SPE allows for clean-up and enrichment to comply py with screening Experimental: ml of sample is fortified with 00 ng ISTD (Glyphosate, 3 C- 5 N, AMPA 3 C) Borate buffer (ph 0) and FMOC (5 mg/ml) is added and reaction takes place over night (> 4 h) Acetic acid (58%) is added d to neutralize sample FMOC derivates are stable for > 48 hours 8

29 Example 3: Robust Method for Glyphosate and AMPA elative Responses Re EIC for Cal_ (0.0 ng/l) + MRM ( > 79.0) Std_0-r00.d x0 3 AMPA 7.0 min Counts > 79.0, >.0 x0 Ratio = 68.5 (8.9 %).8 Counts Acquisition Time (min) Acquisition Time (min) Quantifier: Qualifier: Peak width: 5.4 s Data points: 5 Linear range: 0 ng/l to.0 µg/l 9

30 Example 3: Robust Method for Glyphosate Use of internal standards needed to correct for derivatization yields in different water samples Online-SPE increases the sensitivity to the relevant concentration range (0 to 5 ng/l) for environmental samples Robustness of the method is substantially increased run already > 000 samples on the same chromatographic column with just cartridges AMPA has been found in concentrations up to 000 ng/l, Glyphosate just positive if AMPA concentrations are extremely high 30

31 Review: On-line versus Off-Line Trace Enrichment (SPE) On-Line Increased sensitivity-inject entire sample-analyzed Small Sample Volumes (ml) Increased sample throughput Reduce error-sample handling No evaporation step Off-Line Only a fraction of concentrated sample injected -00 ul from ml extract Large sample volume 00mL-L Extraction and analysis are separate Increase chance of error-manual Concentration step needed No reconstitution Reconstitution Integrated System Standalone Instruments 3 8/9/0

32 Summary and Conclusions The on-line SPE combined with a LC-QQQ system allows a relatively simple, fast and reliable determination of herbicides in the low ng/l range in filtered water samples. The whole system is fully controlled with the MassHunter acquisition software. Adding a more sensitive QQQ system allows for even better sensitivity or compounds which are less weakly ionizable. Good recovery values and reproducibilities can be achieved even in complex samples and for very polar compounds. Online SPE not only increases the sensitivity but adds robustness to the method (e.g. for glyphosate) The chemistry remains the same adding acid to the loading solvent increases the recovery for acidic compounds, adding salt to the sample solution increases the adsorption coefficient for some compounds. The use of online SPE with scanning instruments allows for unknown screening in complex environmental samples. 3 8/9/0

33 I m going in for a sample Thank You! Any Questions? 33 8/9/0